diff --git a/lib/Service/OpenAiAPIService.php b/lib/Service/OpenAiAPIService.php
index 1c4bede7..14208d9d 100644
--- a/lib/Service/OpenAiAPIService.php
+++ b/lib/Service/OpenAiAPIService.php
@@ -496,24 +496,33 @@ public function createChatCompletion(
 				$messages[] = $message;
 			}
 		}
-		if ($userPrompt !== null || $userAudioPromptBase64 !== null) {
-			$message = ['role' => 'user', 'content' => []];
+		if ($userAudioPromptBase64 !== null) {
+			// if there is audio, use the new message format (content is a list of objects)
+			$message = [
+				'role' => 'user',
+				'content' => [
+					[
+						'type' => 'input_audio',
+						'input_audio' => [
+							'data' => $userAudioPromptBase64,
+							'format' => 'mp3',
+						],
+					],
+				],
+			];
 			if ($userPrompt !== null) {
 				$message['content'][] = [
 					'type' => 'text',
 					'text' => $userPrompt,
 				];
 			}
-			if ($userAudioPromptBase64 !== null) {
-				$message['content'][] = [
-					'type' => 'input_audio',
-					'input_audio' => [
-						'data' => $userAudioPromptBase64,
-						'format' => 'mp3',
-					],
-				];
-			}
 			$messages[] = $message;
+		} elseif ($userPrompt !== null) {
+			// if there is only text, use the old message format (content is a string)
+			$messages[] = [
+				'role' => 'user',
+				'content' => $userPrompt,
+			];
 		}
 		if ($toolMessage !== null) {
 			$msgs = json_decode($toolMessage, true);
diff --git a/tests/unit/Providers/OpenAiProviderTest.php b/tests/unit/Providers/OpenAiProviderTest.php
index fe0d2bd5..1d9d694b 100644
--- a/tests/unit/Providers/OpenAiProviderTest.php
+++ b/tests/unit/Providers/OpenAiProviderTest.php
@@ -140,7 +140,7 @@ public function testFreePromptProvider(): void {
 		$options = ['timeout' => Application::OPENAI_DEFAULT_REQUEST_TIMEOUT, 'headers' => ['User-Agent' => Application::USER_AGENT, 'Authorization' => self::AUTHORIZATION_HEADER, 'Content-Type' => 'application/json']];
 		$options['body'] = json_encode([
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
-			'messages' => [['role' => 'user', 'content' => [['type' => 'text', 'text' => $prompt]]]],
+			'messages' => [['role' => 'user', 'content' => $prompt]],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
 			'user' => self::TEST_USER1,
@@ -204,7 +204,7 @@ public function testEmojiProvider(): void {
 		$message = 'Give me an emoji for the following text. Output only the emoji without any other characters.' . "\n\n" . $prompt;
 		$options['body'] = json_encode([
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
-			'messages' => [['role' => 'user', 'content' => [['type' => 'text', 'text' => $message]]]],
+			'messages' => [['role' => 'user', 'content' => $message]],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
 			'user' => self::TEST_USER1,
@@ -269,7 +269,7 @@ public function testHeadlineProvider(): void {
 		$message = 'Give me the headline of the following text in its original language. Do not output the language. Output only the headline without any quotes or additional punctuation.' . "\n\n" . $prompt;
 		$options['body'] = json_encode([
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
-			'messages' => [['role' => 'user', 'content' => [['type' => 'text', 'text' => $message]]]],
+			'messages' => [['role' => 'user', 'content' => $message]],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
 			'user' => self::TEST_USER1,
@@ -334,7 +334,7 @@ public function testChangeToneProvider(): void {
 		$message = "Reformulate the following text in a $toneInput tone in its original language. Output only the reformulation. Here is the text:" . "\n\n" . $textInput . "\n\n" . 'Do not mention the used language in your reformulation. Here is your reformulation in the same language:';
 		$options['body'] = json_encode([
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
-			'messages' => [['role' => 'user', 'content' => [['type' => 'text', 'text' => $message]]]],
+			'messages' => [['role' => 'user', 'content' => $message]],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
 			'user' => self::TEST_USER1,
@@ -402,7 +402,7 @@ public function testSummaryProvider(): void {
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
 			'messages' => [
 				['role' => 'system', 'content' => $systemPrompt],
-				['role' => 'user', 'content' => [['type' => 'text', 'text' => $prompt]]],
+				['role' => 'user', 'content' => $prompt],
 			],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
@@ -469,7 +469,7 @@ public function testProofreadProvider(): void {
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
 			'messages' => [
 				['role' => 'system', 'content' => $systemPrompt],
-				['role' => 'user', 'content' => [['type' => 'text', 'text' => $prompt]]],
+				['role' => 'user', 'content' => $prompt],
 			],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,
@@ -539,7 +539,7 @@ public function testTranslationProvider(): void {
 		$options['body'] = json_encode([
 			'model' => Application::DEFAULT_COMPLETION_MODEL_ID,
 			'messages' => [
-				['role' => 'user', 'content' => [['type' => 'text', 'text' => 'Translate from ' . $fromLang . ' to English (US): ' . $inputText]]],
+				['role' => 'user', 'content' => 'Translate from ' . $fromLang . ' to English (US): ' . $inputText],
 			],
 			'n' => $n,
 			'max_completion_tokens' => Application::DEFAULT_MAX_NUM_OF_TOKENS,